Thursday, February 5, 2015

Just so! How the finch got it's beak.

While some might think the greatest challenge in science is to find an explanation for a particular phenomenon, I would argue that an even greater challenge is to discern from among many reasonable possibilities, which explanation is the correct one. That is, the problem isn’t so much “problems with no solutions” but rather “problems with too many solutions.” A recent trip to Galapagos and a recent paper on zebras have prompted me to ruminate on this topic.

At the most basic level, ADAPTATION is an obvious, and usually safe, one-size-fits-all solution to the problem of understanding variation in organismal traits. Other solutions, such as drift and constraint, are also possible but pale in importance – as I argued in an earlier post. Beyond saying particular differences reflect adaptation, however, we often seek to infer the specific environmental feature driving adaptation. In some cases, this driving feature is obvious: the beaks of Darwin’s finches are the result of adaptation to different food types. In many other cases, the specific force generating natural selection is harder to establish, a point made strongly and cogently by Endler (1986), Wade and Kalisz (1990), and MacColl (2012). Stated another way, a particular trait value in a particular population is almost certainly the result of adaptation – but adaptation to what? Different foods? Different predators? Different parasites? Different abiotic conditions? And which foods, which predators, which parasites, and which abiotic conditions? Thus, the specific selective reason for adaptation of particular organismal traits is often a problem with too many solutions.*

How the zebra got its stripes

Zebras in Krueger National Park, South Africa. Photo A. Hendry.
Classically, zebra stripes were thought to have evolved as an optimal illusion that confuses predators such as lions. This solution is what we all learned as children, and it makes good and obvious sense. No need to look any further. Yet other hypotheses have been suggested. One that has received considerable recent attention is parasite avoidance. Egri et al. (2012) placed similar-shaped but differently-colored models out in nature and found that biting flies were less likely to approach the striped models than the non-striped models. A third hypothesis is that alternating dark and light bands cause differential heating across the skin that generates eddies of air that have a cooling effect. In cases like this, where one problem has multiple solutions, each camp tends to entrench and generate further support for their pet idea rather than stepping back and attempting a test that might formally discriminate among the potential solutions. The paper that partly motivate this post did just that. Larison et al. (2015) examined the relationship among zebra populations between banding patterns and predators (lions), parasites (flies), and temperature. Non-existent correlations for the first two predictors and a strong correlation for the third predictor generates tips the balance in favor of temperature as the driving force behind the evolution of zebra stripes. Of course, I doubt the other hypotheses will die – at least not right away.

From Larison et al.

How the stickleback lost its armor

Marine threespine stickleback colonized freshwater watersheds thousands of times following the retreat of Pleistocene glaciers. Each time they did so, they evolved a dramatic reduction in defensive armor – especially the bony plates on their sides, but also in the size of their pelvis and their dorsal and pelvic spines. Moreover, these evolutionary changes can occur very quickly, such as when humans introduce marine fish into freshwater, eliminate freshwater stickleback thus allowing marine fish to re-invade, or trap marine fish in freshwater. The genetic basis of a number of these changes is well known, but the specific environmental (selective) reason is not. First, the amount of armor in a freshwater population is strongly associated with the resident predators, suggesting that release from the even more intense marine predation is the primary reason for the loss of armor in fresh water. (Even here, uncertainty exists as to which predators – birds, fish, or invertebrates – are the most important.) Second, the amount of armor sometimes correlates strongly with ionic concentrations in fresh water, suggesting that the loss of armor results from limitations in the raw materials needed to build armor. Other ideas abound, including the recent suggestion that armor is too heavy for the low-density medium of fresh water. To date, none of these hypotheses have been strongly excluded from consideration.

Differences in armor plating between marine (top) and freshwater (bottom) stickleback. The image is from Cuvier and modified by D. Kingsley (I found it here)

How the tropics got so speciose

Problems with too many solutions exist not only in evolutionary biology, but also in ecology. For instance, many hypotheses have been suggested for why species richness is higher in the tropics; candidate solutions include increased evolutionary speed (e.g., shorter generation times), fewer disturbances (e.g., a lack of continental glaciers), larger areas provide more opportunities for isolation, and so on. The same explosion of hypotheses attends other ecological phenomena, such as why Atlantic cod populations have not recovered despite 20 years without fishing (e.g., seal predation, Allee effects during breeding, life-history evolution) and why snowshoe hare and lemming populations cycle (e.g., predators, food limitation, stress, life history changes). Interestingly, although these phenomena are “ecological,” many of the proposed solutions are evolutionary.)

One of these hangs on the wall of my office.

How the finch got its beak

This brings me to Galapagos and its finches. More generally, I want to ask how/why beaks evolved. The evolution of this trait was no small thing – bird beaks bear little resemblance to dinosaur teeth. How and why did this change happen? It is surely adaptive, but what was the specific selective force driving the change? Perhaps the most widely accepted solution is that beaks dramatically reduce weight for a flying animal, just as do their hollow bones. (Yet bats have teeth, and some beaks are rather heavy.) Another solution is that beaks were particularly well-suited for eating seeds. (Yet seeds were around for a long, long time before beaks evolved, and many animals that do not have beaks eat seeds.) Yet another is that beaks are so adaptable that a lineage with beaks would be more likely to persist and diversify – the “key innovation” solution. I recently had an epiphany stemming from personal experience that leads me to suggest yet another solution to the evolution of beaks. To illustrate where this epiphany started, I must digress for a moment.

The (really big) beak of the finch - a large-beaked ground finch (Geospiza magnirostris) from Santa Cruz Island. Photo A.Hendry
I take very good care of my teeth. I brush – hard and long – twice a day. I floss religiously once a day – vigorously. And it seemed to work. I don’t think I had a single cavity for the first 20+ years of my life – not one. Yet it has recently all gone to pot. Now I probably have 15 fillings, most of them in the last 5 years. The funny thing is that I tend to notice incipient cavities when I am in the Galapagos, because I get closer to the mirror there than I do at home. At home, I have a counter between me and the mirror and so I never see my teeth closely: not so in Galapagos, where counters aren’t present and sinks are tiny. Last year I noticed some brown smudges on my teeth, which turned out – on my return – to indeed be cavities. This year I noticed some more, and while stewing from the immediate frustration that resulted, I walked out to where Kiyoko, Diana, and Luis were discussing finch beaks. Bang – epiphany.

Tooth decay can’t be stopped, even in modern humans, who are aware of the problem and combat it with the best technologies/tools and the greatest incentives. (How the hell did the dental industry convince employers to offer such good insurance when the same is not true for vision?). Indeed, before these technologies, tools, and incentives, humans suffered horribly from tooth decay. Pretty much any forensic anthropology display at any museum shows numerous instances of horrible abscesses, worn teeth, and missing teeth. Yet even these pre-modern humans knew that tooth decay was a bad thing (some cleaned their teeth) and tried to prevent/fix it. Coincidentally, here at McGill, we have evidence of the earliest dental intervention in history, in an Egyptian mummy housed in our Redpath Museum. We also have display a display of teeth that a street-corner dentist had removed, sort of a “Bad teeth? I can get rid of ‘em” advertisement.

The same problem must attend non-human animals, which do not have the same foresight nor technologies. Such animals should have frequent dental problems that can cause death through systemic infection or starvation. Thus, tooth decay must surely have reduced the fitness of many animals in nature.** Several arguments might be leveled against this hypothesis. First, non-human animals might not live long enough to get tooth decay – but some do live a long time and tooth decay can occur early in life. Second, tooth decay in humans might be somewhat modern problem that evolved after the development of processed sugars – but tooth decay was also prevalent before such sugars. Third, tooth decay might well predate processed sugars but might be due to our high-carbohydrate diet – but other animals also have such diets. Fourth, the ancestor of birds likely replaced its teeth as do most lizards - but this still represents a cost.
A captive lion with so many tooth problems that it "went off its feed" until given false teeth
I suppose you long ago saw where I was going with all this. BIRDS DON’T GET TOOTH DECAY. I propose that bird beaks evolved – at least in part – for this reason. Of course, I am not saying that avoidance of tooth decay was what started beak evolution – but it would certainly be a benefit that could accelerate the process once it started. It is also true that the fitness costs of tooth decay in the ancestor of birds might not have been that dramatic, since dinosaurs seemingly replaced their teeth gradually over their life. Yet this represents a cost of its own; and the signatures of tooth decay have been found in dinosaurs. So, why not?

In closing, I had better make clear that I am just having fun here by posing a “just-so” story for the evolution of bird beaks. This admission would seem to invite the criticism heaped on “the adaptationist programme” ever since Gould and Lewontin’s classic paper The Spandrels of SanMarco and the Panglossian Paradigm. Yet the truth is that just-so stories are always the starting point of any scientific explanation. By this I mean that one can’t possibly test an idea until one first has the idea, and ideas are always just-so stories until they are tested. Now we just need someone to recreate the transition between teeth and beaks so that we can turn our just-so stories into that’s-why stories.


* Too many solutions to a problem might simply reflect the fact that the solution is multifarious; perhaps adaptation was simultaneously driven by multiple causal factors (predators AND parasites AND temperature) and the fact that different factors might be important in different locations (predators HERE ionic concentrations THERE and buoyancy OVER THERE).

** After writing this post, I looked up “tooth decay in wild animals” on the internet and most places asserted that they don’t get tooth decay – for a variety of reasons. Yet these were just assertions by pundits, not serious analyses by scientists. Then I found “A Literature Review of Dental Pathology and Aging byDental Means in Nondomestic Animals:Part II.” In this paper P.T. Robinson reports that gross examination of herbivores and carnivores from the wild review few cavities but criticizes these counts as biased. (I would add that, if my hypothesis is correct, they might well have died before a hunter could shoot them.) Robinson also reported that more detailed analyses reveal high levels in older individuals of some species, including 25% in old capuchin moneys. Regardless, animals are certainly known to have many dental problems of various sorts that would have the same effect. (Of course, beaks can also break.)

A lion with a broken tooth - from the Mara Predator Project.


  1. Toothed birds didn't get tooth decay. Nor did other dinosaurs, crocodiles or lizards.

    Mammals have a bizarre system where we only get two sets of teeth (milk and adult), which means that our adult teeth suffer greatly from wear and bacterial action over time. Diapsids, like dinosaurs lizards and crocs, all have continual tooth replacement. So they'd lose a tooth and grow a new one. Which means no individual tooth lasted long enough to rot.

  2. Also, it's worth noting that beaks have evolved *a lot* of times. Turtles and birds are the most famous living examples, but within the group of dinosaurs that birds evolved from, beaks evolved minimally four times. They also evolved in other groups of dinosaurs, as well as some bizarre crocodile relatives from the Triassic ( Also, in a group related to mammals called Dicynodonts.

    In all of those groups, it's worth noting that the evolution of the beak is associated with a transition to herbivory. Carnivorous birds are all secondarily carnivorous, and I don't know of any examples where beaks evolved in a carnivorous clade.

  3. Thanks for these details. Of course, my post was mostly just fun. However, teeth can cause problems unrelated to decay, most commonly broken teeth. And tooth replacement is costly itself. Although, like I said, I am just having fun here.



A 25-year quest for the Holy Grail of evolutionary biology

When I started my postdoc in 1998, I think it is safe to say that the Holy Grail (or maybe Rosetta Stone) for many evolutionary biologists w...